The behavior of alpha2-plasmin inhibitor in
fibrinolytic states.
N Aoki, … , M Matsuda, K Tachiya
J Clin Invest.
1977;
60(2)
:361-369.
https://doi.org/10.1172/JCI108784
.
Human plasma alpha2-plasmin inhibitor in fibrinolytic states was studied using
immunochemical methods and radioiodinated plasminogen. The concentration and activity
of plasma alpha2-plasmin inhibitor decreased when urokinase was added to plasma in vitro
or infused intravenously in man. The decrease was associated with the appearance of
plasmin-alpha2-plasmin inhibitor complex which subsequently disappeared from the
circulation in a short time. A decrease of other major inhibitors, such as
alpha2-macroglobulin and alpha1-antitrypsin, was not observed when the amount of urokinase
added or infused was relatively small, and conversion of plasminogen to plasmin was not
extensive. The formation of plasmin-alpha2-macroglobulin complex was observed only
when plasma plasminogen was activated with a larger amount of urokinase, and after most
of the alpha2-plasmin inhibitor was consumed by forming complexes with plasmin. The
formation of plasmin-alpha1-antitrypsin complex was not observed even in the highly
activated plasma unless exogenous plasmin was added to the plasma. alpha2-Plasmin
inhibitor was the only inhibitor of which the concentration in plasma was significantly
decreased in patients with disseminated intravascular coagulation and fibrinolysis among
the major plasmin inhibitors in plasma. The most reactive inhibitor for regulating plasma
fibrinolysis very likely is alpha2-plasmin inhibitor.
Research Article
Find the latest version:
The
Behavior of a2-Plasmin
Inhibitor
in
Fibrinolytic
States
NOBuo AoKi, MASAAKI
MOROI, MICHIO
MATSUDA, and KAZUKO TACHIYA
From the Department of Medicine and the Institute of Hematology, Jichi Medical School, Minamikawachi-Machi, Tochigi-ken 329-04,Japan
A
B S
T R A C T
Human plasma
a2-plasmin inhibitor in
fibrinolytic states was studied using immunochemical
methods and radioiodinated plasminogen. The
con-centration
and activity of plasma
a2-plasmin
in-hibitor decreased when urokinase was added to plasma
in
vitro or infused intravenously in man. The
de-crease was
associated with the appearance
of
plasmin-a2-plasmin inhibitor complex which
sub-sequently disappeared from the circulation
ina
short
time. A
decrease of other major inhibitors, such as
a2-macroglobulin and
a1-antitrypsin, was not observed
when the amount of urokinase added or infused was
relatively small, and conversion
of plasminogen to
plasmin
was not extensive.The formation of
plasmin-a2-macroglobulin complex was
observed only
when
plasma
plasminogen
wasactivated with a
larger amount
of
urokinase,
and
after
mostof the a2-plasmin
in-hibitor was consumed by
forming
complexes with
plasmin. The formation
of
plasmin-al-antitrypsin
complex
was notobserved
even inthe
highly
acti-vated
plasma unless exogenous
plasmin
wasadded
to the
plasma. a2-Plasmin inhibitor
wasthe
only
inhibitor
of which the concentration
inplasma
wassignificantly decreased
inpatients with disseminated
intravascular coagulation and
fibrinolysis among the
major
plasmin inhibitors
inplasma. The
most reactiveinhibitor
for regulating plasma fibrinolysis very
likely
is
a2-plasmin
inhibitor.
INTRODUCTION
A
plasma
fraction
isknown which inhibits
plasmin-ogen activator-induced clot
lysis
efficiently
but
pos-Part ofthisdata was presentedat the Intemational Society
ofHematology Meeting,8September1976.Theresearch was
carried out according to the provisions of the Declaration of Helsinki, and informed consent was obtained from each
one ofthepatients.
Receivedfor publication23September1976andinrevised form 1April 1977.
sesses
only a small part
of the whole capacity of
plasma to neutralize plasmin activity (1, 2). This
fraction
iscalled
antiactivator, since it wasfound
tobe
different from the major known plasmin inhibitors
such asa2-macroglobulin (a2-M)l and a1-antitrypsin
(a,-AT),
both which strongly inhibit plasmin on incubation but
exert
little effect on activator-induced clot lysis (1, 3).
This antiactivator protein was purified, and its
proper-ties were
investigated (4). The antiactivator
wasulti-mately
found to be primarily a plasmin inhibitor
be-longing to a2-globulin,
and the
term.
a2-plasmin
inhibitor (a2-PI) or a2-proteinase inhibitor is
con-sidered to
be appropriate. a2-PI
wasshown
tobe
dif-ferent
from a2-M,
CIinactivator,
inter-a-trypsin
in-hibitor,
a1-antichymotrypsin,
a,-AT,
and antithrombin
III
(4). a2-PI
inhibits plasmin
almost
instantaneously,
and the inhibition of
plasminogen
activationby
theinhibitor was found to be mainly
due to the very
rapid inactivation of plasmin
formed (4).
Inthis paper,
we
consider that this unique
biological property of
a2-PI has
animportant role
inregulating
fibrinolysis
invivo.
METHODS
Plasma. Human blood was collected from antecubital veins into a 0.1-volof3.8%trisodiumcitrateandwas
centri-fuged at 2,000 g at tip for 20 min to prepare platelet-poorplasma.
Preparation of purifiedplasminogen and plasmin. Plas-minogen was prepared from human freshplasma according
tothe method ofBrockwayand Castellino (5). Plasminogen fraction2, whichwaseluted after fraction 1,wasused for all the studies.Plasmin wasprepared byactivatingplasminogen
with urokinase-coupled Sepharose (Pharmacia Fine Chem-icals, Div. ofPharmacia, Inc., Piscataway, N.J.) according
tothe methodpreviously described(4).Activity of
plasmino-gen and plasmin wasassayed accordingtothe caseinolytic
method of Robbins and Summaria (6). Protein
concentra-'Abbreviations used in this paper:a2-M,
a2-macroglobulin;
a2-PI, a2-plasmin inhibitor; a1-AT,
a,-antitrypsin;
DIC, dis-seminated intravascular coagulation.tion of purified plasminogen was determined by measuring the absorbance of the plasminogen solution (pH 7.4) at 280 nm and by converting absorbance to protein concentration usingE19m= 17.0 for purified plasminogen (7).
Urokinasepreparations. Urokinase preparations used for intravenous infusion were obtained from Green CrossCorp.,
Osaka and from Mochida Pharmaceutical Co., Tokyo. Ac-tivity was expressed by intemational units based on the standard preparations supplied by the World Health Organ-ization. The unit is nearly equivalent to the Committee of Thrombolytic Agents unit.
Assayfor plasma plasminogenand fibrinogen. Plasmin-ogen in plasma was measured after destruction of
anti-plasmin by acidification (8). All plasmas to be assayed
were first acidified by the addition of the equal volume of 0.166 N HCI, allowed to stand for 15 min, reneutralized with 0.166 N NaOH, and then subjected to the caseinolytic method (6). The results were expressed in Remmert and Cohen casein units (9). The fibrinogen content of plasma wasassayed by the method of Ratnoff and Menzie (10).
Immunochemicalmethods. Rabbit antisera against human
a,-AT
anda2-M were obtained from Behring-Werke AG,Mar-burg/Lahn, WestGermany. Rabbit antiserum against human a2-PI was prepared by immunizing rabbits with a purified preparation of the inhibitor (4). About1mgof protein in 1 ml
of 0.05 M Tris-HCl, pH 7.4, containing 0.15 M NaCl
was mixed completely with an equal volume of Freund's complete adjuvant and was injected in equally divided volumes intracutaneously into each footpad, shoulder, and
thigh of a rabbit. 3 wk later, the injection was repeated
with 0.5 mg protein. Immune sera were harvested2-5 wk
afterthesecondinjection,andwereabsorbedwith an a2-PI-free plasmafraction instep 4 of the purification procedures
ofthe inhibitor described previously (4). The fraction had
been passed through the plasminogen-coupled Sepharose
column repeatedlytoremove a2-PI completely. 1 vol of this plasma fraction, absorbancy of whichwas 17 at 280nm,was
added to 40 vol of antiserum, incubated at 37°C for 15
min, and stored at 4°C for 48h. The resulting precipitate was removed by centrifugation anddiscarded.Thesupernate
was storedina small aliquot with 0.1% ofsodium azide as a preservative. Antiserum thus obtained was shown to be
monospecific by double immunodiffusion and
immuno-electrophoresis aspreviouslydescribed(4). The IgG fraction
ofantisera wasobtainedbyDEAE-cellulosechromatography
(11). Rabbitantiserum against humanplasminogen was
pre-pared by immunizing rabbits with a purified preparation of plasminogen.Immunizationprocedurewasessentially the
same asthe one usedforantiserum against a2-PI described
above.About10mg of proteinwasused for the first injection and 1.5 mg for the second injection. Immune sera were ab-sorbed with plasminogen-free human plasma which had
passed through a column oflysine-Sepharose (4). 1 vol of
fivefold-diluted plasminogen-freeplasma was added to9vol
ofantiserum, andthe precipitate formedwasremoved. Theconcentration ofa2-PI in human plasma or serum was
determined bythe one-dimensional method of Laurell (12) using1% antiserum inagarose (0.8%inbarbitalbuffer,pH 8.6, ionic strength 0.05). 10
p1
of antigen solution (purifiedinhibitor, plasma, or serum) was applied to the electro-phoresis which wasrunwithaconstant current of1 mA/cm
width for 7h. Dilutions ofknown amounts of the purified
inhibitor were made with Tris-buffered saline (Tris 0.05 M, NaCl 0.15 M, pH 7.4) containing 3% wt/vol bovine serum
albumin and used for calibration. There was a linear
rela-tionship between peak heights and amounts of antigen ap-plied (Fig. 1). The purified inhibitor used for calibration
was the one which had the highest specific activity, 1,933
U/ml per
A280
(4), in a series of preparations and appeared as a single band on sodium dodecyl sulfate polyacrylamide gel electrophoresis. The protein concentrations of the puri-fied inhibitor were determined by measuring the absorbance of the inhibitor solution (pH 7.4) at 280 nm and converting absorbance to protein concentration usingE""m
= 7.03 for pure inhibitor (4). The extinction coefficient,EF^3m,
was de-rived from the measurement of absorbance of the solution of known weight of the purified inhibitor, which had been extensively dialyzed to remove salts, lyophilized, and then dried in a vacuum at 1 10°C over P205 for 4 days.The concentrations of a2-M and
a,-AT
in human plasma or serum were determined by the single radial immunodif-fusion technique using partigen plates (Behring-Werke AG).Antigen-antibody-crossed electrophoresis was performed essentially according to Laurell (13). 30 ul of antigen solution was placed in a transverse slit (1.5 x 10 mm) in the agarose layer (0.8% agarose in barbital buffer, pH 8.6, ionic strength 0.05, and 1.8 mm thickness) on a glass slide (2.5 x 7.5 cm). The electrophoresis was performed for about 2Y2 h with a constant current of 2 mA/cm width. After-wards the gel slip was transferred to a second glass plate (5 x 7.5 cm). The remainder of the second plate was then covered with a layer of agarose containing 1% antiserum against a2-PI, 0.4% antiserum against plasminogen, or 2% antiserum against a,-AT, separately. The second electrophore-sis was performed by applying the electric current (1 mA/ cm width) at an angle of 900 to the direction of the first electrophoretic separation. The time of the second run was 10h. The buffer used for preparation of agarose gel and electrophoretic run was barbital buffer (pH 8.6, ionic strength 0.05). The plates were washed with saline to remove non-precipitated protein and dried. The immune precipitates were then stained with amido black.
'a2M'
antiplasmin activity. Antiplasmin activity ofplasma was estimated by the modified method of the fibrino-lytic assay described previously (1). 0.1 ml of 30
,ug/ml
(0.52 casein U/ml) plasmin in 25% glycerol was mixed and incubated at 37°C for 30 min with 0.1 ml of the test plasma which had been diluted 20-fold with barbital-buffered saline (5.2 mM barbital 0.14 M NaCl, pH 7.4). The mixture was then transferred to an ice-waterbath, diluted with 0.6 ml of cold barbital-buffered saline, and left for 3 min. Subse-quently, 0.1 ml of fibrinogen solution followed by 0.1 ml ofCONCE
2 4 6 8 IO 12
NTRATION OF INHIBITOR (mg/lOOml)
FIGUREI Areferencecurve for measuring theconcebtration
ofa2-PI by the one-dimensional method of Laurell (12). A linear relationship between concentration of purified
aL2-PI
and length ofprecipitate (peak heights) was obtained. For details see the text.362
N.Aoki,
M. Moroi,M.Matsuda, and
K.Tachiya
LU
c
t:
u LU cy-CL
LL
20 U/mlbovine thrombin (Parke, Davis & Company, Detroit,
Mich.) was added, and the solution was transferred to a 370C water bath, and the clot lysis time was recorded.
Fibrinogensolution used was 2% Cohn's fraction I of human plasma (Green Cross Corp., Osaka) which had been treated with lysine-Sepharose to remove plasminogen (14) and
contained1.6g/100 ml of fibrinogen. The control was run by replacing the test plasma with buffered saline. The differ-encebetween the control measurement and the test run, ex-pressed in fibrinolytic plasmin units as defined previously (1), divided by the volume (0.1 ml) of the test
sample,
multiplied by the dilution factor 20 represents the anti-plasmin units per milliliter. The determinations of the activityofplasma samples from 12 healthy adults revealed that the mean ±2 SD of normal values are 83±42 U/ml. The antiplasmin activity measured by this method is referred to
as a2-M antiplasmin activity, since the method measures mainly the antiplasmin activity of a2-M and is insensitive
to the change of a2-PI activity (1-3). Insensitivity of the method to the change of a2-PI may be explained by the fact that the method measures nearly total capacity of antiplasmin activity of whole plasma (1), and a2-PI contributes only 2.5%
onaverage of the total antiplasmin activity (see Discussion). Assay of
ag-PI
activity. Activity of a2-PI in plasma or serum was assayed by the method based on the inhibitoryactivity of plasma or serum to activator-induced clot lysis (1, 2). Activator-induced clot lysis is principally inhibited bya2-PI but also bya2-M toasmall degree (2). When plas-minogen-free plasma prepared by lysine-Sepharose (4)was
chromatographedonSephadex G-200 (PharmaciaFine Chem-icals, Div. of Pharmacia, Inc.) as described previously (2),
and each fraction ofthe effluentwasanalyzedforinhibitory
activity to urokinase-induced clot lysis (2), the a2-M frac-tionwhich emerged inthe first protein peak accounted for approximately 10-20% of the total activity of plasma
de-pendingona2-Mconcentration ineachplasma sampleused. The inhibitory activity exerted by the a2-M fraction was
diminishedtoapproximately halfby freezing and thawingthe
fraction, whereas inhibitory activity ofa2-PI fraction which
emerged
later inthechromatography was notaffected byafreezing-thawing procedure.Thiswasalsoshownbythestudy
usingpurified a2-M (15) and purified a2-PI. Inhibitoractivity
of8mg/mla2-Mwas 120U/ml whenassayed bythe present
method for a2-PI activity. The activity was diminished to
55 U/ml by freezing and thawingthe same sampleof a2-M. However, a freezing-thawing procedure did not affect
ap-preciablythe inhibitor activity ofpurified a2-PI.
Therefore,
freezing and thawing the test sample before the assay
de-creasestheactivityby approximately 5-10lo and makes the assay method more specific to a2-PI. The decrease of
in-hibitory activity of a2-Mtoproteolytic enzymes(trypsin and
chymotrypsin) by freezingandthawing
procedures
has beenreportedalsoby others (16). Sincethe presence of
plasmino-gen in the test sample influences the assay method and giveserroneouslylowervaluesofinhibitor activity,
plasmin-ogen in the test sample was removed before the assay
by
lysine-Sepharose. 1 ml ofthe test plasma or serum, once
frozen and thawed, was applied on a column of lysine-Sepharose (1 ml)
equilibrated
withbarbital-buffered saline. The plasma wasallowed togo intothecolumn slowly andwasfollowed with barbital-bufferedsaline. The first5mlof
theeffluent,which containedmore than 95% of theprotein
applied,wascollected. This fivefold-dilutedsamplewasused for the assay. An aliquot of the sample was diluted in the
test tube (10x75 mm) to 0.7 ml with cold
barbital-buffered saline and was mixed with 0.1 ml of2% Cohn's fraction Iofhumanplasma which contained 1.6g/100mlof
fibrinogenand 1.2 casein U/mlofplasminogen. Themixture
was cooled in an ice-water bath. To this mixture were added 0.1 ml of 600 CTA U/ml urokinase (Mochida Pharmaceutical Co., Tokyo), and 0.1 ml of 20 U/ml bovine thrombin to make a final clot of 1 ml. The tube was quickly shaken to mix the reagents well and placed in a 37°C water bath. A stopwatch was started at the time of addition of thrombin. Soon there formed a clot in which many small air bubbles werefonned and trapped. The clot lysis was made noticeable by the sudden rise of the air bubbles to the upper surfaces. In another set of experiments, similar clot lysis tests without test sample were performed with various concentrations of urokinase. A standard curve was constructed by plotting on double-logarithmic paper the clot lysis time obtained by various concentrations of urokinase against units of urokinase in a 1-ml clot. A straight line resulted over the range of 6-60 U urokinase/ml clot. Using this standard curve, clot lysis time obtained with a sample containing the inhibitor wasconverted to a value of urokinase activity. There was an inverse relationship between the amount of the inhibitor pres-ent and resulting residual urokinase activity, which is linear in the region from 60 urokinase U (control with urokinase and no inhibitor) to 20 U/ml clot (urokinase plus various amounts of inhibitor). The inhibitor units per milliliter plasma or serum were calculated as the difference between the measurement of urokinase activity in the presence and in the absence of the test sample, expressed as units per milliliter clot divided by the volume (milliliters) of the test sample introduced into the assay system and multiplied by 5 (the dilution factor derived from the initial treatment of plasma or serum with lysine-Sepharose). The mean ±SD of the activities of a2-PI in plasma from 29 healthy Japanese adults measured by this method were 1,224+374 U/ml.
Specificity
of the clot lysis method fora2-PI.
The IgG fraction of rabbit antiserum to a2-PI or a2-M was added to lysine- Sepharose-treated and frozen-thawed plasma, in-cubated at 37°C for 1 h and further at 4°C for 24 h. The resulting precipitates were removed by centrifugation. The residual inhibitor activity in the supernate was measured by the clot lysis method. 2, 1, 0.5, and 0.25 mg of IgG fraction ofanti-a2-PI antiserum neutralized 95, 90, 82, and 70% of the inhibitor activity of 1 ml plasma, respectively. The similar amounts of IgG fraction of anti-a2-M antiserum did not ap-preciably affect the inhibitor activity. This indicates that the assay method is specific to some extent for a2-PI, but the ef-fect of any other proteinase inhibitor in plasma on the assay can not be excluded. Thrombin preparation used in the assay was crude material and might have been loaded with non-thrombin biologic activities. However, when purified thrombin prepared by the method of Lundblad (17) was used instead of crude thrombin, it gave the same results as thoseobtained with crude thrombin, suggesting that any con-taminant in the crude thrombin preparation did not affect the assay.Radioiodination of plasminogen. Plasminogen was radio-iodinated by modification of the method of Greenwood et al. (18). To 13 mg of plasminogen in 6 ml of Tris-saline (0.05 M Tris-HCl, pH 7.4, 0.15 M NaCl) were added 0.15 mg of Nal in 0.15 ml of 0.05 M phosphate buffer (pH 7.4) and 80 ,uCi of 125I-Na solution (The Radiochemical Centre, Amersham, England) while stirring in an ice water bath. Immediately thereafter, 0.4 mg of chloramine-T (Tokyo Kasei Co., Ltd., Tokyo) in 0.1 ml of phosphate buffer was added, and the mixture was stirred for 1 min. Then 0.1 ml of Na2S202 solution (4 mg/ml phosphate buffer) was added to stop the reaction. Separation of
1251-plasminogen
from the reaction mixture was carried out by gel filtration through a column ofSephadex G-50. Elution was initially carried outwith 0.6 ml ofthe mixture of NaI (0.4 mg) and KI (2 mg), followed byTris-saline. Specific activity ofplasminogenbefore and after radioiodination were 23.1 and 20.2 U/mg protein, re-spectively.
Distribution of
'251-plasminogen
orplasmininurokinase-activated plasma. 20 ul of radioiodinated plasminogen (0.91 mg/ml) was added to 2 ml plasma,
yielding
plasma containing9,ugradioiodinated plasminogen/ml.This plasma was activated with various amounts of urokinase and wassubjected to electrophoresis in agarose in the same way as
the firstelectrophoresisof
antigen-antibody-crossed
electro-phoresis. Immediately after the completion of electro-phoresis, the gel was cut transversely into 3-mm slices. Each slice was put into a counting vial, and radioactivitywas counted by Auto Well GammaSystem, Aloka JDC-752 (Aloka Co., Tokyo).
Statistical analysis. Statistical analysis was performed by Student'stanalysisforpairedorunpaired sampleswhere applicable. Forunpaired samples, Snedecor and Cochran's modification was applied (19). A P value>0.05 was
con-sidered to representastatisticallynonsignificantchange.The two-tailedtest wasused unless otherwise indicated.
Patients. Effects of intravenous infusion of urokinase on
plasmaplasmininhibitorswerestudiedon sevenpatients who received urokinase therapy within 72h after an attack of cerebral infarction. 240,000 U of urokinasewasadministered to each patient except foronepatientwho received 120,000U.
Patients having disseminated intravascular coagulation (DIC) with fibrinolysis were studied for the concentrations ofplasmin inhibitors. The diagnosis of DIC was based on
clinical observations as well as laboratory findings such as
increasedfibrin/fibrinogendegradation products
(16-640,ug/
ml, normal< 4,ug/ml [20]), decreasedplasma antithrombin III activity (10-80% of normal standard, normal range 90-110% [21]),prolonged prothrombin time (> 15
s,.normal
12-13 s [22]), decreased plasminogen (0.2-1.4 casein U/ml, normal range 1.4-2.0 casein U/ml), decreased fibrinogen(20-200 mg/100 ml,normalrange 150-400mg/100 ml),and
decreasedplatelet count (<
150,000/,A,
normalrange150,000-400,000/,ul)
measured by Coulter counter (Coulter Elec-tronics Inc., Hialeah, Fla.). These values were obtained by the repeated assays duringthe course of the diseases rather thanbyadeterminationon a single occasion. The gradual orsudden decline of platelet count, fibrinogen, plasminogen, and antithrombin IIIactivity, togetherwithfurther
prolonga-tionof prothrombintimeand persistent presenceor asteady increase of fibrin/fibrinogen degradation products, was most suggestive of DIC.Underlyingdiseases were carcinoma,
acuteleukemia,and septicemia.
RESULTS
Development of the complexes of plasmin and
in-hibitors in
urokinase-activated
plasma in vitro.
Plasma
wasincubated
with various amounts ofuro-kinase, and subsequently subjected
toantigen-anti-body-crossed electrophoresis.
Controlplasma without
addition of urokinase contained
a2-PI as a singlecomponent
(Fig. 2A). When plasma was activated withurokinase,
there appeared a newcomponent
which
possessed
theantigenicity
of
a2-PI with a slowermobility
inthe electrophoretic field
asindicated
by arrow 1 (Fig. 2A). The newcomponent became
moreprominent,
and the a2-PIcomponent
itself wasdimin-ished
when the amount of urokinase wasincreased,
as seen in the
bottom
figure of Fig. 2A. The newcomponent
disappeared
after theplasma
wasabsorbed withantiplasminogen
antiserum,
indicating
that it isa
complex
betweena2-PI
andplasmin
generated
by
urokinase. When
antigen-antibody-crossed
electro-phoresis
was carried outusing
antiplasminogen
anti-serum, the appearance of another newcomponent
with a faster
mobility
was found in theplasma
activated with
high
concentrations of urokinase(>
100U/ml)
asindicated
by
arrow 2 inFig.
2B. Thiscomponent became less
prominent
when theplasma
was absorbed
before
electrophoresis
withanti-a2-M
antiserum,
indicating
that the component is thecom-plex
ofa2-M
andplasmin.
It isnoteworthy
that thecomplex
ofa2-M
andplasmin
did notdevelop
when the concentration ofurokinase
was as low as 25 U/ml(Fig. 2B),
whereas the formation of thecomplex
ofa2-PI
andplasmin
was observed(Fig.
2A).
FIGURE 2 Formation of the complex of the plasma inhibi-tors and plasmin in plasma activated with urokinase in
vitro, demonstrated by antigen-antibody-crossed
electro-phoresis. To each 1 ml of plasma was added 2.5 and 50
pA of urokinase solution (10,000 U/ml) separately, and the mixtures were incubatedat37°Cfor 2 hand further atroom
temperature foranadditional 12 h. They were subsequently
subjectedtoantigen-antibody crossed electrophoresis using antiseraagainst a2-PI (A) orplasminogen (B). The direction ofthe first electrophoresis was from the left to the right.The amounts (units)ofurokinase added to 1 ml plasma are
indi-cated between the panels. Arrows indicate the complex of
plasmin and inhibitors. Arrow 1, the complex of a2-PI and plasmin. Arrow 2, the complex of a2-M and plasmin.
364
N. Aoki, M. Moroi,
M.Matsuda, and
K.Tachiya
RR',
gmmm
:.un
When these urokinase-activated plasma samples were
analyzed by
the crossedelectrophoresis
using antiserum againsta,-AT,
noformation
of the complexof
a1-ATand plasmin
wasobserved,
even in plasmaactivated with high
concentrations of urokinase, such as 500U/ml plasma.
In these plasma samples, nearly all plasminogen was converted to plasmin, and most of the a2-PIformed
complexes with plasmin (Fig. 2). The formation of the complex ofa,-AT
and plasmin was observed only after the addition of exogenousplasmin
tothis highlyactivated
plasma.Plasma containing radioiodinated plasminogen was
activated
withurokinase, and
thedistribution of
radioactivity in plasma fractions was analyzed by
electrophoresis.
Theresults
are shown in Fig. 3. Whenplasma
wasactivated
with 25 U of urokinase/ml plasma, thepeak
ofradioactivity shifted
to the location (slice 5)corresponding
tothe peak of a2-PI-plasmin
complex indicated by
arrow 1 in Fig. 2A (Fig. 3A). Thepeak of a2-PI-plasmin complex became
prominent with concurrent decrease of the original plasminogenpeak (slice
3) whenamounts
of urokinase used
wereincreased
(Fig. 3B). With increasingurokinase,
theredeveloped another
newpeak of radioactivity (slice
9) whichcorresponds
to the location of thepeak
indi-300
X200
A t
u 100 r
0
30
p
In
B
10
CONTROL UK 25UNITS
2 4 6 8 10 12 14 SLICE NUMBERS
0o-K125UNITS
nI
.1
U 2 4 6 8 10 12 14
SLICE NUMBERS
FIGuRE 3 Distribution ofradioactivity in plasma fractions
after activation of plasma containing radioiodinated
plas-minogen with urokinase.Toeach 1mlofplasmawasadded
2.5, 12.5, and 50 ,ul of urokinase solution (10,000 U/ml)
separately, andthemixtureswere incubated at 37°C for2h and further at room temperature for additional 12h. They
were subsequently subjected to electrophoresis in agarose,
and the distribution of radioactivity was analyzed. The amounts ofurokinase addedto 1 mlplasmaareindicatedon each figure. Control was run withoutaddition of urokinase. Slices of agarose plate are numbered from the origins wherethe samples wereapplied. Details of the methodare describedin thetext. UK, urokinase.
>o[
-NoNDEPLETION
< IOO X2MDEPLETED
2 4 6 8 10 12 14
SLICE NUMBERS
FIGURE 4 Depletion of plasmin-a2-M complex by
anti-serumagainsta2-M. 0.1 ml ofantiserum was addedto 1 ml
plasmawhichcontained radioiodinatedplasminogenandhad been incubated with 125 U/ml urokinase as in Fig. 3. The mixture was then incubated at 37°C for 30 min and further at40C for additional 12h. The resulting precipitates were
removed by centrifugation andthe supernatewas subjected
toelectrophoresis. Thecondition of the electrophoresis and the subsequent treatment were the same as in Fig. 3. The
control(nondepletion)was runby replacing rabbitantiserum
with unimmunized rabbit serum.
cated by
arrow 2 in Fig. 2B (Fig. 3B). This new peak is most likelyattributable
to the formation of the a2-M-plasmin conmplex, in that the peak was destroyed by the treatment of the plasma with a2-M anti-serum before electrophoresis (Fig. 4).Development of
thecomplexes
ofplasmin
and
inhibitors
inplasma
of patients receiving urokinase intravenously. For therapeutic purposes 240,000 or 120,000 Uof
urokinase together with 5,000 U of heparin wasinfused intravenously for
2h to sevendifferent
patients withcerebral
infarction.
Immedi-ately before the infusion, and
at 0, 3, 6and
18hafter the
terminationof the
infusion, blood samples
were
withdrawn
from the patientsand subjected
toanalyses by
antigen-antibody-crossed electrophoresis (Fig. 5). Thecomplex of
a2-PIand plasmin
wasob-served
prominently
inthesamples withdrawn
immedi-ately
after the infusion (Fig. 5). Thecomplex
became less prominent at3h, andnocomplex
was found at6h after the infuision (Fig. 5). The antigen concentration,as well as the activity of a2-PI,
decreased
concur-rently
with the appearance of thecomplex of
a2-PIand
plasmin.
The decrease of activity was morere-markable than
thatof
antigenconcentration at the timeimmediately after
the infusion (Fig. 5). Thecomplex
of
plasmin and
a2-M ora,-AT
asdescribed
in invitro
studies
mentioned above
was notfound
at any time.Decrease
of
a2-PI after theintravenous
infusionof
urokinase. To continue our
studies
in these patients,concentrations (antigen) of the inhibitors,
a2-PI,
a2-M, and
a,-AT,
in plasma were measuredimmuno-chemically
before and after the intravenous infusion (2 h) of 240,000 U of urokinase together with 5,000 Uof heparin. Inhibitory activity of
a2-PI,
a2-Manti-plasmin activity, anti-plasminogen, and fibrinogen were also
assayed.
The concentrations and the activity ofE a
(a.
-11.
FIGuRE 5 Formationand disappearance of the complex of a2-PI and plasminin plasma ofthe patientsunder urokinasetreatment,demonstratedbyantigen-antibody-crossed electrophoresis
using antiserum againsta2-PI. Plasma sampleswerewithdrawn from thepatientsbefore (a),
im-mediately after (b), and6hafter (c) the infusion of urokinase.Patients AandBreceived 120,000 and 240,000 U ofurokinase, respectively. Arrows indicate the complex ofa2-PI and plasmin. The concentration (mgantigen/100 ml) and the activity (U/ml shown in parentheses) ofa2-PI are indicated under each figure.
a2-PI were
significantly decreased by the infusion of
urokinase
and had not yet returned
tothe
original
values when
examined at 18 h after the termination of
infusion (Tables
Iand II). On the
contrary, the
con-centrations
ofa2-M and a1-AT were not changed
signif-icantly by the infusion. Some significant increase of
a2-M
antiplasmin activity was noticed at the
timeim-mediately after the infusion (Table II). Plasminogen
TABLE I
ChangesintheConcentrations ofInhibitors inPlasma aftera 2-hUrokinaseInfusion
Timeafterthe start ofinfusion, h
Inhibitors 0 2 20
mg/100ml
a2-PI 4.78±+1.27* 3.42+0.52 3.02±0.29
P<0.02t P<0.02
a2-M 191+22 191+19 198+26
NS NS
a,-AT
244+40 248+41 261+24NS NS
* Mean+SD
(n.=
6).t P valuesreferto the difference in concentrations insamples
at2or 20 hfrom those in samplesat 0 time.
decreased significantly after urokinase infusion,
whereas fibrinogen remained unchanged (Table III).
Decrease of
atrPI
in DIC with fibrinolysis.Con-centrations
of
a2-PI, a2-M,and
a,-AT
inplasma
weremeasured immunochemically in patients having DIC with
fibrinolysis, and
compared
withnormal
values.The
mean +2 SDof
the concentrationsof
a2-PI inplasma
from 25healthy
Japaneseadults
were 5.98 ±+1.76 mg/100 ml. Thereported values of
the mean±2
SD of concentrations
ofa2-M and
a,-AT
inplasma
TABLE II
ChangesinInhibitorActivitiesofPlasma after
a 2-h Urokinase Infusion
Timeafter thestartof infusion, h
Inhibitors 0 2 20
U/mi
a2-PI 1,461+298* 588±206 911+244
P<0.001t P<0.005
a2-M 103+22
136±
18 95+22antiplasmins P < 0.05 NS
* Mean+SD
(a2-PI,
n =6; a2-M, n=5).t P values refer to the difference in activities in samples at 2or 20 h from those in samples at 0 time.
366
N.Aoki, M. Moroi, M.Matsuda,
andK. TachiyaTABLE III
Changes intheConcentrationsof Plasminogen and FibrinogeninPlasma aftera2-h
Urokinase Infusion
Time afterthestart ofinfusion, h
0 2 20
Plasminogen,
casein U/ml 1.41±0.26* 1.06±0.22 1.05±0.3
P<0.0005t P<0.05
Fibrinogen,
mg/lOO
ml 394±123 337±91 346±86NS NS
*Mean±SD(n =5).
t Pvalues refertothe differencein concentrations in
samples
at2or20 hfrom thosein
samples
at0time.One-tailedtest.from
160healthy Japanese
adults measured by single
radial
immunodiffusion are
210+100and
200+100mg/100
ml, respectively (23). The concentrations of
a2-PI
in mostof the patients
werefound
tobe lower
than the
normal range, whereas values of a2-M
were inthe
normal range except for
onecase, and those
ofa,-AT
werehigher
inmany
casesthan the normal
range (Fig. 6). Many of the values of a2-M,
eventhough
inthe
normal range,
werewell below the
meanlevel of normals, thus
implying that there had been
some
reduction
inthe
level
inthose
cases. Nocom-plex of
plasmin
with any
oneof the inhibitors
wasdetected when these
patients' plasmas
wereanalyzed
by
antigen-antibody-crossed
electrophoresis.
a2-PI
0
C!4
E
6
2
Ot2-M
800
600 300
-0
*0 0
-40-0
*t
O1
200-
1001-OL
P<0.001
0
*.0
I
---a___
4001
2001-o0 NS
FIGURE 6 Concentrations of a2-PI, a2-M, and al-AT in
plasma ofthe patients with DIC, measuredbythe
immuno-chemical method. Themean concentrationofeach inhibitor is indicatedbyashorthorizontalline. Themeans +2 SDof
normal values forhealthyadults are indicated bysolidand
interrupted horizontallines,respectively.Pvalues refertothe difference in concentrations in DIC from those in normal.
DISCUSSION
The concentration of a2-PI in normal plasma was
foundto
be approximately 5-7 mg/100 ml, which is calculated
to
be _0.7-1.0 ,uM on the basis of mol wt 67,000.
The molar concentrations of a2-M (mol wt 820,000) and
al-AT (mol wt 54,000) in normal
plasma are estimatedon the
average at 2.4 and 37
,uM,
respectively, while
their average concentrations are 210 and 200 mg/100
ml, respectively (23). Consequently, it may be
con-cluded
onthe assumption
of one-to-one molar reaction
of
inhibitor and enzyme that a2-PI contributes only
2.5% on average of the
total capacity of antiplasmin
activity of the
whole plasma. However, a2-PI has the
strongest
affinity for plasmin among these inhibitors.
It had been suggested that the strong affinity of a2-PI
for
plasmin plays an important role in the regulation
of
fibrinolysis. To test this possibility, the behavior of
a2-PI
infibrinolytic states was explored in the present
study.
When plasminogen in
plasma was activated in vitro
by
arelatively low concentration of urokinase, there
appeared readily
aplasmin-a2-PI complex. However,
no
complex of plasmin-a2-M or
a,-AT
was
observed.
The formation of the plasmin-a2-M complex was
ob-served
only when plasma
wasactivated with a larger
amount
of urokinase and after
mostof the a2-PI
was
consumed by forming complexes with plasmin.
The formation of
plasmin-al-AT
complex was not
ob-served
evenwhen
all of the plasminogen in plasma
was
converted to plasmin unless exogenous plasmin
was
added.
When a relatively low
dose of urokinase was infused
intravenously
topatients, the consistent
findings were
the
decrease of the concentration and the activity of
a2-PI
and the
concomitantformation of the
plasmin-a2-PI
complex. The decrease of the activity of
a2-PI
was moreremarkable than that of the
con-centration atthe
timeof the
existenceof the
complex
in
blood. This
canbe
explained by the fact that the
concentration
measured
by the immunochemical
method includes the enzyme-inhibitor complex which
has
lost the activity.
In contrast
to a2-PI,
concentrationsof a2-M and
a,-AT
were not
significantly changed, and
nocomplex
of plasmin-a2-M
or-a,-AT
wasobserved.
A
significant increase of a2-M antiplasmin activity
observed
immediately
after the
infusion
might
be
at-tributable
toantiplasmin
activity of
heparin-anti-thrombin
III(24), because there
was nosignificant
increase
of a2-M antigen
concentrationthroughout
thestudy, and
nosignificant
increaseof
a2-M
antiplasmin
activity
wasobserved
18hafter
the
terminationof
infusion
when noresidual effect of heparin
wasde-tected.
Arnesen and
Fagerhol
studied effects of urokinase
fusion
ona2-M,a,-AT,
andantithrombin III inplasma (25). The doses of urokinase thatthey
used were far greater thanthose
used
in the presentstudy.
They infusednearly twofold
the amount ofurokinase
ascompared
to ours in 10 min as aloading
dose and continued infusion of a maintenance dose for18h.
Their
hourly
maintenance dose isnearly
equivalent
to
the total
amountof
urokinase infused
into each patient inthe
presentstudy.
In spiteof their large
dose, the decrease of
a2-M aftercompletion
ofin-fusion
wasonly
20% on average of thepreinfusion
values.
There
was nochange of
plasma
antithrombin IIIlevel, and the
concentrationof a,-AT
was evenincreased. Nilehn and
Ganrotactivated practically
allplasminogen by infusing
alarge dose of
strepto-kinase
and observed the decrease of
a2-M to 50% ofthe initial
values
(26). In the presentstudy,
the de-creaseof
plasminogen level averaged
25% ofthe
pre-infusion values, and
nodecrease
of a2-M wasob-served. These data, together
withthose
obtained by
others (27, 28), indicate that a2-M
is oneof the
majorantiplasmins which
reactwithplasmin
generated
dur-ingfibrinolytic therapy.
The presentstudy provides,
furthermore, information
on anewly discovered
in-hibitor
inaddition
toa2-M.
The
presentstudy
suggests that a2-PI is the mostreactive
plasmin
inhibitor
inplasma, and whenever
plasminogen
activationtakes
place,
a2-PI isthe
firstinhibitor
which reacts withplasmin formed. Only
whenthe
excess amountof
plasmin
isformed, and when
most a2-PI
has been consumed by forming complexes
with
plasmin,
then a2-M willbecome
theprincipal
inhibitor
toplasmin. This
view issupported by the
finding that,
amongthe
majorplasmin inhibitors
inplasma,
a2-PI wasthe
only inhibitor of
whichthe
concentration was
significantly decreased
in patients with DIC. However, nocomplex
ofplasmin-a2-PI
wasobserved
inplasma of these
patients, atleast
by
themethod presently employed.
This isexplained by
therapid
disappearance
ofplasmin-a2-PI complex from
the
circulating blood, which,
in the presentstudy,
was
indeed shown after the infusion of urokinase.
Theplasmin-a2-PI complex formed after intravenous
in-fusion of 240,000
Uof urokinase
disappeared
from the
circulating blood
within 6h. Thecomplex
wasproba-bly removed by the
clearance
mechanismof the
retic-uloendothelial
systeminorgansespecially
intheliver,
since the
complex
isformed by
astable covalent
bond (4), and its
dissociation
can notreadily
occur.a2-M might
also
have reacted with plasmin in these DIC patients, inasmuch as almost all values of a2-M werebelow
the mean valuealthough
they were in the normal range. However, this possibility was notsta-tistically
proved in the present study (P> 0.1). The cause ofthe increase ofa,-AT
observed
in these DICpatients
was notknown,
but it isinteresting
to notethat
a,-AT
concentration inplasma
wassteadily
increasing
during prolonged
infusion of urokinase(25).
Because the
molecular weight of
the complex ofa2-PI
and thelight
chain ofplasmin
waspreviously
reported
as84,000
(4)
andthe molecular weight
of theheavy
chain wasnearly 67,000 (4),
the molecularweight
ofplasmin-a2-PI complex
must be close to150,000.
Theplasmin-a2-PI complex might be
thesame as the
complex reported by
Collen et al.(29).
They reported that, during
the activation ofplasmin-ogenin
plasma
with urokinaseinvitro, thereappeared
the
complex
ofplasmin
and ahitherto
unidentifiedantiplasmin.
The molecularweight
of thiscomplex
wasestimated
by gel
filtration to be in the region of 130,000-170,000, which is the same region ofthe
estimate of molecular
weight
ofplasmin-a2-PI
com-plex.
Miillertz(30)
also reported thaturokinase-activated
plasma contained,
ata low concentration ofurokinase,
a component withplasmin
+plasminogen
antigenicity which was eluted with
nearly
7S protein fromSephadex
G-200 and had8-mobility by
gel
electrophoresis.
This component appears from itsphys-ical properties to be identphys-ical with
plasmin-a2-PI
complex.
Inaddition to these
observations,
plasma inhibitors offibrinolysis
which possess properties similar to those of a2-PI were recently described (31-35). The questionas to whetherornota2-PI is identical with any one
of
these inhibitors needs further
investigation, including
elucidation of the
physicochemical
andimmunochem-ical properties of each inhibitor.
Crossed
immunoelectrophoresis
of someplasma
samples
using anti-a2-PI antisera gave double immuno-precipitate lines as seen in patient B in Fig. 5. Development ofdouble immunoprecipitate lines
might be due to the molecular
heterogeneity
of antigen as suggested by Laurell (36), and the antiseramight
contain
antibodies
against more than one of the determinants. Molecularheterogeneities
of a2-M andal-AT
havealready
been revealed by crossedim-munoelectrophoresis
(37). Another possibility that the antiserum is not specific and reacts with two differentmolecules
which possess the same electrophoreticmobility
cannotbeexcluded,
althoughimmunoelectro-phoresis
anddouble immunodiffusion
have not been able todemonstrate
this possibility (4).ACKNOWLE DGMENTS
Theauthors would like to thank Dr. M. Yoshida and Dr. A.
Ueki of the Department of Neurology at the Jichi Medical
School fortheir generous cooperation. Acknowledgment is
also due to Dr. W. H. Seegers, Wayne State University, for
hiscritical reading.
This work was supported bya grantfrom the Ministry of Education of the Govemment of Japan.
REFERENCES
1. Aoki, N., and K. N. von Kaulla. 1971. Human serum plasminogen antiactivator: its distinction from anti-plasmin. Am. J. Physiol. 220: 1137-1145.
2. Aoki, N., and T. Kawano. 1972. Inhibition of plasminogen activators by naturally occurring inhibitorsinman. Am.J.
Physiol. 223: 1334-1337.
3. Aoki, N., and M. Moroi. 1974. Distinction of serum
in-hibitor of activator-induced clot lysis from
a,-anti-trypsin. Proc. Soc. Exp. Btol. Med. 146: 567-570. 4. Moroi, M.,and N. Aoki. 1976. Isolation and
characteriza-tionof a2-plasmin inhibitor from human plasma. A novel proteinaseinhibitor which inhibits activator-induced clot lysis. J. Biol. Chem. 251: 5956-5965.
5. Brockway, W. J., and F. J. Castellino. 1972. Measure-ment of the binding of antifibrinolytic amino acids to variousplasminogens. Arch. Biochem. Biophys. 151: 194-199.
6. Robbins, K. C., and L. Summaria. 1970. Human plasmin-ogen andplasmin. Methods Enzymol. 19: 184-199. 7. Barlow, G. H., L. Summaria, and K. C. Robbins. 1969.
Molecular weight studies on human plasminogen and plasmin at the microgram level. J. Biol. Chem. 244: 1138-1141.
8. Alkjaersig, N.,A. P. Fletcher, and S. Sherry. 1959. The mechanism of clot dissolution by plasmin.
J.
Clin. Invest. 38: 1086-1095.9. Remmert, L. F., and P. P. Cohen. 1949. Partial puri-ficationand properties ofaproteolytic enzyme of human serum.J. Biol. Chem. 181: 431-448.
10. Ratnoff, 0. D., and C. Menzie. 1951. A new method for the determination of fibrinogen in small samples of plasma. J. Lab. Clin. Med.37: 316-320.
11. Rowe, D. S., and J. L. Fahey. 1965. A new class of human immunoglobulins. II. Normal serum IgD. J. Exp. Med.
121: 185-199.
12. Laurell, C. B. 1966. Quantitative estimation of proteins
by electrophoresis in agarosegel containing antibodies. Anal. Biochem. 15: 45-52.
13. Laurell, C. B. 1965. Antigen-antibody crossed electro-phoresis. Anal.Biochem. 10: 358-361.
14. Matsuda, M., S. Iwanaga, and S. Nakamura. 1972. A
simple, large scale method for preparation of
plasmino-gen-free fibrinogen.Thromb. Res. 1: 619-630.
15. Iwamoto, M.,andY.Abiko. 1970. Plasminogen-plasmin
system. Preparation of a2-macroglobulin antiplasmin from human plasma. Biochim. Biophys. Acta. 214: 402-410.
16. Lebreton de Vonne, T., H. Mouray, G. Berthillier, and
R.Got. 1972.Influencedu vieillissement des a2-macro-globulines de lapin sur la formation du complexe a2-macroglobuline-enzymes. Biochim. Biophys. Acta. 257:
365-371.
17. Lundblad, R. L. 1971. A rapid method for the
purifica-tion of bovine thrombin and the inhibition of the puri-fied enzyme with phenylmethylsulfonyl fluoride.
Bio-chemistry. 10:2501-2506.
18. Greenwood, F. C., W. M. Hunter, and J. S. Glover. 1963. The preparation of 1311-labelled human growth hormone of high specific radioactivity. Biochem. J. 89:
114-123.
19. Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. IowaState University Press, Ames, Iowa. 6th
edition. 114-116.
20. Merskey, C., G. J. Kleiner, and A. J. Johnson. 1966. Quantitative estimation of split products of fibrinogen
in human serum, relation to diagnosis and treatment.
Blood. 28: 1-18.
21. Monkhouse, F. C. 1963. The influence of inorganic salts
on plasma antithrombin activity. Thromb. Diath. Haemorrh.9: 387-394.
22. Tocantins, L. M. 1964. Estimation of prothrombin
(One-stage method of Quick).In Blood Coagulation,
Hemor-rhage and Thrombosis. L. M. Tocantins and L. A. Kazal, editors. Grune & Stratton, Inc., New York. 148-150.
23. Matsuda, T. 1967. Coagulation studies on strokes. Nihon Rinsho. 34: 72-78.
24. Highsmith,R.F., andR. D.Rosenberg.1974.The
inhibi-tion of humanplasmin by human antithrombin-heparin cofactor.J. Biol. Chem. 249:4335-4338.
25. Arnesen, H., and M. K. Fagerhol. 1972.c2-Macroglobulin, a1-antitrypsin, and antithrombin III in plasma and serum during fibrinolytic therapy with urokinase. Scand. J. Clin. Lab. Invest. 29:259-263.
26. Nilehn, J-E., and P. 0. Ganrot. 1967. Plasmin, plasmin inhibitors anddegradation products of fibrinogenin
hu-man serum during and after intravenous infusion of streptokinase. Scand. J. Clin. Lab. Invest.20: 113-121.
27. Fischer, M. 1971. Influence of thrombolytic therapy on human antiplasmins. Thromb. Diath. Haemorrh. 47
(Suppl.): 153-157.
28. Spottl, F., and F. Holzknecht. 1970. The influence of
inhibitors of plasmin and plasminogen activation onthe
streptokinase-induced fibrinolytic state. Thromb. Diath. Haemorrh.24: 100-112.
29. Collen, D., F. De Cock, and M. Verstraete. 1975.
Immunochemical distinction between antiplasmin and
alpha1-antitrypsin.Thromb.Res. 7: 245-249.
30. Mullertz, S. 1974.Different molecular forms of plasmin-ogen and plasmin produced by urokinase in human
plasma and their relation to protease inhibitors and lysis of fibrinogen and fibrin. Biochem. J. 143: 273-283. 31. Hedner,U. 1973.Studieson aninhibitor of plasminogen
activation in human serum. Thromb. Diath. Haemorrh. 30: 414-424.
32. Mui,P. T.K.,H. L.James, and P.Ganguly. 1975. Isola-tion and properties of a low molecular weight
anti-plasmin of human blood platelets and serum. Br. J.
Haematol.29: 627-637.
33. Gallimore, M. J. 1975. Serum inhibitors offibrinolysis.
Br.J. Haematol.31: 217-231.
34. Hedner, U., and D. Collen. 1976. Immunochemical
dis-tinction between the inhibitors of plasminogen activa-tion and antiplasmin in human plasma. Thromb. Res. 8: 875-879.
35. Edy, J., F. De Cock, and D. Collen. 1976. Inhibition ofplasmin by normal and antiplasmin-depleted human plasma. Thromb. Res. 8: 513-518.
36. Laurell, C. B. 1972.
Electroimmunoassay.
Scand.J.Clin.Lab.Invest. 29(Suppl. 124):21-37.
37. Ganrot, P. 0. 1972. Crossed immunoelectrophoresis.
Scand.J.Clin.Lab. Invest. 29(Suppl. 124): 39-47.